There's an old saying made popular by Mark Twain about 'lies, damned lies, and statistics'.
As the CEO of a company that works with research data I usually take some issue with that quote - but something the UK Office for National Statistics (ONS) has just done left me shaking my head in disbelief. Last week they reported on new "experimental research" they have undertaken - trying to measure how many lesbians, gay men and bisexual people (LGB) there are in the UK.
My company Out Now works in the LGBT market so this is research we are most pleased to see happening, especially by an organisation as significant as the ONS.
Sadly though the research the ONS have produced can only best be described using that most British of expressions - as being 'not fit for purpose'.
According to the ONS statistics published last week a mere 1.5% of Brits are able to be identify themselves as being one of the three letters of the LGB community in the UK.
How many LGBT youth groups, suicide prevention lines, anti-bullying groups or gay health initiatives are in line to see their funding slashed because the 'official' count of LGBT people in the UK according to the ONS has now just been revised downwards - by more than 75%?
Given that most other studies over the years have shown us numbers closer to 6% or even 7% and above (like Yankelovich MONITOR in the US in 1992 and the peer-reviewed University of Indiana Human National Survey of Sexual Health and Behavior study of 2010) for the number of LGB people living in the USA we wondered what on earth can be happening in Britain to make LGB people so damned scarce. Well it turns out they're not. They are there - and they can even be discerned in the ONS data if you just know where to look.
You see what the ONS did that was so different from those reputable US studies I mentioned was the ONS took little to no trouble at all to establish respondent trust.
Whereas in the US Yankelovich research respondents got to know the person interviewing them over several weeks on more than one personal visit, ONS respondents were phoned at home by a stranger, who knew where they lived; or were asked to reveal their sexuality to a personal interviewer at home - knowing all members of the same household over the age of 16 would shortly be asked that same personal question.
To put that into some kind of context - Out Now undertakes the world's largest global LGBT research initiative - the LGBT2020 study www.lgbt2020.com - and that research showed that in 2012 many UK LGBT people are not open about their sexuality to all the people in their own lives that they know rather well.
Only 52% of Brits feel able to be out with all their work colleagues. For the ONS to expect all their respondents to simply reveal their sexuality to a stranger in a survey is just naive. The 'International LGBT2020 Homophobia Report' shows that, of the more than 2,500 UK respondents sampled in 2012 for the Out Now LGBT2020 study, only 56% of respondents said they were able to be out to all their family members.
Being reluctant makes sense. 53% of respondents witnessed homophobia in their own workplace last year. 49% of LGBT2020 respondents were verbally harassed in the previous twelve months due to their sexual orientation or gender identity. 6% were harassed by neighbours and 15% experienced personal harassment in their own workplaces because they are LGBT.
So in all probability - to use a statistical phrase - it seems fair to assume many respondents the ONS interviewed would feel less than thrilled about sharing something so significant about themselves with a complete stranger.
We know of course that the ONS published data has to be wrong - and not just because it is so significantly lower than the US research.
The ONS report itself includes a data table showing that up to 4% of their sample can be found in the "Don't know / refusal" category for the question on their own sexuality.
Of course some in this world are yet to decide for sure what allocated label they want to ascribe to their own sexuality - but for the ONS to just ignore that 4% elephant in the room of their study is not only poor research, it is dangerous.
We are amazed it did not strike anyone at the ONS as perhaps a bit 'odd' that this many people might not know their own sexual orientation?
If we are therefore to conclude that these respondents are in fact refusing to divulge their sexuality - a not unreasonable assumption we contend, given the clumsy way in which respondents are expected to 'out' themselves to a total stranger - then the only honest headline that the research the ONS released last week could really appear under is: "Up to 4% of UK people prefer not to divulge their sexuality to a stranger when asked to for an ONS research study".
If you think like we do that this is a dangerous survey masquerading as credible data you might want to let the Director General at the ONS know. We wrote to him here and you can email his office with your own thoughts at email@example.com - I am sure they might look into it if enough people questioned their very questionable research methodology.
Saying nothing means that this extremely dodgy data might be allowed to stand in the UK - and I fear such an outcome would most definitely make Mr Twain right on this particular occasion.